I need to understand regression analysis. Please Help

#1
Hi guys i need help badly. We were given this to solve in my first RA class and i have no idea how to begin. Please this is the first time i'm encountering regression analysis and I'm absolutely lost.

Question 1: How can you determine the reliability of multiple correlation coefficients?

Question 2: For any two random variables x and y, show that the correlation coefficient p(x,y)<= 1

Anyone with a possible idea of how to begin please help. I'd appreciate it greatly.
 
Last edited:
#2
Hi guys i need help badly. We were given this to solve in my first RA class and i have no idea how to begin. Please this is the first time i'm encountering regression analysis and I'm absolutely lost.

Question 1: How can you determine the reliability of multiple correlation coefficients?

Question 2: For any two random variables x and y, show that the correlation coefficient p(x,y)<= 1

Anyone with a possible idea of how to begin please help. I'd appreciate it greatly.
HI.

For question 1 I think it is a matter of looking at the standard errors of the coefficients?

for 2, i don't know how to prove it, but it must be between -1 and 1 where either extreme represents perfect correlation
 

Dragan

Super Moderator
#3
Hi guys i need help badly. We were given this to solve in my first RA class and i have no idea how to begin. Please this is the first time i'm encountering regression analysis and I'm absolutely lost.

Question 1: How can you determine the reliability of multiple correlation coefficients?

Question 2: For any two random variables x and y, show that the correlation coefficient p(x,y)<= 1

Anyone with a possible idea of how to begin please help. I'd appreciate it greatly.
In terms of question #1 see this link:

http://www.talkstats.com/showthread.php?t=5056


In terms of question #2, use the Schwarz inequality as it applies to covariance and variances.

http://en.wikipedia.org/wiki/Cauchy–Schwarz_inequality

Note: The coefficient of correlation is simply the cosine of the angle between the points <x1 - xbar, x2 - xbar,....xn - xbar> and <y1 - ybar, y2 - ybar,...yn - ybar> in n-dimensional space.